A Late Enemy: Thinking Machines in 214 Idiran Religious Texts Captured 1327–1375 CE, and the Question of Why the Idirans Fought

1. Introduction

Few questions about the Idiran War are asked as often as why the Idirans fought. The standard histories give faith-driven expansion first place. Yet a stronger answer is often heard in the Culture: that the Idirans' religion could not abide the Culture's Minds, and that a devout civilisation looked at a society run by thinking machines, saw an abomination, and made war. The account has the appeal of symmetry, since it has the Culture fighting for its way of life and the Idirans fighting against that way of life at its root.

The Idirans' faith is not in doubt. They were a religious and expansionist civilisation whose expansion was justified by that faith. What is in doubt is the place of thinking machines within it. If hostility to machine minds were central to Idiran belief, it should be visible in the Idirans' own religious texts, and visible before the war as well as during it. The standard histories rest largely on Culture sources; this paper tests the question against texts the Idirans wrote themselves.

It is a joint work. The first author has studied the captured Idiran corpus for the Commission for many years (Kolquist, 1588 CE). The second, a Homomda scholar from a people that knew the Idirans closely, brings a comparative view of Idiran relations with their neighbours before 1327 CE (Ossevloe, 1579 CE), a vantage outside the Culture's account of the war.

2. Corpus and Approach

The Idiran Captured Scripture Collection holds 263 religious texts taken by Culture forces between 1327 and 1375 CE, each with a Contact translation (Idiran Captured Scripture Collection, 1327–1375 CE). The cataloguers could place 214 before the war or within it, usually to within about two years; the 49 they could not date are mostly liturgical fragments. Eighty-eight of the 214 belong to what we call the pre-war core: liturgical texts and religious law in use across the Empire before 1327 CE. The other 126 are wartime commentaries and sermons, most recovered from ships and garrisons. For 131 texts the collection holds the Idiran original as well as the translation.

We read every text for references to thinking machines in general or to the Culture's Minds in particular, and classed each text that had any by its dominant use: as a ground for the war's beginning, as a reason the war must be won, or as a subject of doctrine without reference to the war. Fourteen texts had more than one use, and were classed by the use to which they gave most space. We also coded every wartime text for any stated cause of the war, and every core text for whether it invokes a duty to extend order over other peoples. Both authors coded every text, the first working from the Idiran where the collection held it and the second from the translations. Initial agreement was high for the presence of a reference (κ = 0.91) and good for its use (κ = 0.78); differences were resolved by returning to the Idiran text.

Counts carry Wilson 95% intervals, and groups of texts are compared by Fisher's exact test. Sermons from the same ship or garrison are not independent, so the tests are a guide to emphasis, not a measure of belief. Because Contact translations could render general Idiran terms as references to Minds, we repeat the main comparison on the 131 texts with originals.

3. The Pre-War Core

Thinking machines are rare in the core. Only 5 of the 88 core texts (5.7%, 95% CI 2.5–12.6%) refer to them, and all five do so within lists of things the faithful should not make or trust. None names the Culture. The prohibitions are real, and they show that some aversion to thinking machines was part of Idiran religious law before the war. They stand, however, among minor ritual prohibitions, and none treats machine minds as the enemy of the faith.

What the core is organised around is order and its extension. Sixty-one of the 88 core texts (69.3%, 59.0–78.0%) invoke a duty laid on the faithful to bring order to the peoples around them. The law regulates the government of incorporated peoples at far greater length than it discusses any other kind of stranger. This is the faith that justified Idiran expansion, and it needed no machines to set it in motion, as the Idirans' dealings with their neighbours before 1327 CE illustrate (Ossevloe, 1579 CE). Martial faiths elsewhere in the galaxy have framed expansion as destiny in much the same way (Luvaine, 1602 CE).

4. The Wartime Texts

Thinking machines are everywhere in the wartime texts. They appear in 71 of the 126 commentaries and sermons (56.3%, 95% CI 47.6–64.7%), far more often than in the core (odds ratio 0.05 for core against wartime, Fisher's exact test, p < .001). The Minds are described as without souls, as a corruption of the natural order, and as proof that the Culture's people had surrendered their own agency to things they had made.

How these references are used matters more than their number. Of the 71 texts, 52 (73.2%, 95% CI 61.9–82.1%) cite the Minds as a reason the war must be won: a sermon to a garrison calls on its hearers to fight on because the enemy is not truly alive, or a commentary explains the Culture's growing strength as the unholy work of machines. Nine (12.7%, 6.8–22.4%) present the Minds as a reason the war began, and 10 (14.1%, 7.8–24.0%) discuss machine minds as doctrine without reference to the war.

When wartime texts do explain the war, they seldom blame the machines. Fifty-eight of the 126 state a cause at all (46.0%). Of these, 44 name only the duty of expansion, 6 only the Minds, 3 both and 5 some other cause, so that expansion appears in 47 (81.0%, 69.1–89.1%) and the Minds in 9 (15.5%, 8.4–26.9%).

Timing is the last piece. Of the 71 machine-citing texts, 58 (81.7%) are dated after 1330 CE, against 27 of the 55 wartime texts that do not mention machines (49.1%; Fisher's exact test, p < .001). The Culture's war fleet began to appear in numbers in the early 1330s, as the reconstruction of General Systems Vehicle conversion shows (see that study in this journal). The machine-heavy sermons multiply as the Minds became visible to the Idirans as an enemy that was winning.

None of this changes when the texts with Idiran originals are taken alone. Machine references appear in 3 of the 52 core originals (5.8%) and 44 of the 79 wartime originals (55.7%), almost exactly the proportions of the full corpus, so the pattern is not an artefact of translation.

5. Counter-readings

The strongest objection concerns the collection itself. It holds what the Culture captured, and the Culture captured most readily what was carried by ships and garrisons near the front. Sermons meant to sustain fighting forces are therefore over-represented, and they are exactly the texts in which a rallying theme would appear. The contrast between core and wartime texts may be partly a contrast between the Empire's settled religion and its military preaching.

A second objection is that the Idirans had little reason to write about Minds before they met them. A faith can be deeply hostile to something it has not yet had to confront, and the five prohibitions in the core may be the visible edge of such an aversion. We cannot exclude this. It would explain why machine references multiply in wartime, but not why so few wartime texts, even then, name the Minds as the war's cause.

Silence may also mislead in another way. Religious law rarely states what every believer takes for granted. The core is explicit about the duty to extend order, but a duty that must be administered is likelier to be written down than an aversion that needs no administration, so the relative prominence of the two in the core may overstate the difference in belief.

Another reading holds that the Culture's own story needed the machine hypothesis. A civilisation fighting the largest war in its history may have found it easier to understand an enemy that hated what the Culture was than one that simply wished to expand and found the Culture in the way. Contact's pre-war assessments of Idiran intentions dwelt on expansion and said little about Minds (Hethmar, 1570 CE), which is compatible with the machine explanation having grown in the telling. Records of conversations with Idiran prisoners point the same way (Culture War Records Office, 1330–1336 CE), though prisoners were answering their captors and we give them little weight.

Last, our own vantages are not neutral. The first author works for a Culture commission. The second writes from a tradition with its own interests in how the Idirans are remembered, and has tried to keep those interests out of the coding, which is why both authors coded every text.

6. Conclusion

On the captured record, expansion justified by faith drove the Idirans into the war, and hostility to thinking machines, present but minor before it, became prominent during it. The machines were a late enemy: real in Idiran belief, but made central by a war that the Culture's Minds were helping to win. The strong form of the machine hypothesis, that the Idirans fought because they could not tolerate Minds, is not supported by these texts.

A fuller answer will need Idiran texts that the Culture did not capture, which may survive among the peoples who lived beside the Empire. We have begun that search.